Image Classification
PyTorch
timm
Transformers
text-rendering
text-to-image
chinese
cjk
glyph
clip
vit
deberta-v2
ids
quality-assurance
reward-model
Instructions to use ATH-MaaS/TextVerifier-CJK-CLIP with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- timm
How to use ATH-MaaS/TextVerifier-CJK-CLIP with timm:
import timm model = timm.create_model("hf-hub:ATH-MaaS/TextVerifier-CJK-CLIP", pretrained=True) - Transformers
How to use ATH-MaaS/TextVerifier-CJK-CLIP with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-classification", model="ATH-MaaS/TextVerifier-CJK-CLIP") pipe("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/hub/parrots.png")# pip install -U transformers accelerate # Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("ATH-MaaS/TextVerifier-CJK-CLIP", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download prototype_metadata.json from ATH-MaaS/TextVerifier-CJK-CLIP: direct link, hf CLI and curl.
- Browser
- Download file 534 Bytes
-
https://huggingface.co/ATH-MaaS/TextVerifier-CJK-CLIP/resolve/main/prototype_metadata.json
- Command line
-
hf download hf://ATH-MaaS/TextVerifier-CJK-CLIP/prototype_metadata.json
-
curl -L -o prototype_metadata.json https://huggingface.co/ATH-MaaS/TextVerifier-CJK-CLIP/resolve/main/prototype_metadata.json
534 Bytes
| { | |
| "schema_version": 1, | |
| "architecture": "TextVerifier Stage-3, ViT-L/16 image encoder, IDS-aware dual encoder", | |
| "image_size": 128, | |
| "embedding_dim": 1024, | |
| "prototype_shape": [ | |
| 8189, | |
| 10, | |
| 1024 | |
| ], | |
| "num_characters": 8189, | |
| "K": 10, | |
| "threshold": 0.02152947708964348, | |
| "error_score_rule": "1 - max_j cosine(image_embedding, prototypes[target_character, j])", | |
| "prediction_rule": "render_error if error_score > threshold", | |
| "prototype_norm_min": 0.9999998211860657, | |
| "prototype_norm_max": 1.0000001192092896 | |
| } | |